homebrewnlp

Clashluke's group workspace

Group: tied-moe-modulo

3287

252

3495

3497

3491

2631

3498

3496

3494

3497

3490

3494

3148

1-13

of 13

Timestamps visible

2022-10-15 18:44:35

[ 95873/262144] Loss: 0.685 - Accuracy: 0.789 | LearningRate: 0.00512 | StepTime: 40.424603s - Rate: 1,251,445.1 Tokens/s

2022-10-15 18:45:15

[ 95877/262144] Loss: 0.678 - Accuracy: 0.791 | LearningRate: 0.00512 | StepTime: 40.421875s - Rate: 1,250,867.6 Tokens/s

2022-10-15 18:45:55

[ 95881/262144] Loss: 0.681 - Accuracy: 0.790 | LearningRate: 0.00512 | StepTime: 40.405253s - Rate: 1,250,291.0 Tokens/s

2022-10-15 18:46:37

[ 95885/262144] Loss: 0.682 - Accuracy: 0.789 | LearningRate: 0.00512 | StepTime: 40.430225s - Rate: 1,249,714.5 Tokens/s

2022-10-15 18:47:18

[ 95889/262144] Loss: 0.677 - Accuracy: 0.791 | LearningRate: 0.00512 | StepTime: 40.423816s - Rate: 1,249,138.8 Tokens/s

2022-10-15 18:47:58

[ 95893/262144] Loss: 0.689 - Accuracy: 0.787 | LearningRate: 0.00512 | StepTime: 40.425039s - Rate: 1,248,563.6 Tokens/s

2022-10-15 18:48:38

[ 95897/262144] Loss: 0.681 - Accuracy: 0.789 | LearningRate: 0.00512 | StepTime: 40.452166s - Rate: 1,247,988.6 Tokens/s

2022-10-15 18:49:18

[ 95901/262144] Loss: 0.687 - Accuracy: 0.788 | LearningRate: 0.00512 | StepTime: 40.425692s - Rate: 1,247,414.6 Tokens/s

2022-10-15 18:50:00

[ 95905/262144] Loss: 0.671 - Accuracy: 0.793 | LearningRate: 0.00512 | StepTime: 40.423059s - Rate: 1,246,841.1 Tokens/s

2022-10-15 18:50:40

[ 95909/262144] Loss: 0.675 - Accuracy: 0.792 | LearningRate: 0.00512 | StepTime: 40.405904s - Rate: 1,246,268.4 Tokens/s

2022-10-15 18:51:20

[ 95913/262144] Loss: 0.694 - Accuracy: 0.786 | LearningRate: 0.00512 | StepTime: 40.423269s - Rate: 1,245,696.1 Tokens/s

2022-10-15 18:52:00

[ 95917/262144] Loss: 0.684 - Accuracy: 0.789 | LearningRate: 0.00512 | StepTime: 40.424149s - Rate: 1,245,124.4 Tokens/s